Hello and welcome back to Cyber Aires. Microsoft Azure Administrator A Z 103 course. I'm your instructor, Bill Carlson. And this is Episode 27 Virtual Machine considerations.
In today's episode, we're gonna discuss a couple of things that we need to consider before we actually launch, particularly our first production. Bm.
We're gonna discuss the concept of regions and how that affects the availability of service is in the pricing of those service's and azure.
We'll also discuss the different families of virtual machines. So the different things that they're geared to accomplish for us, and also some changes to the way virtual machine discs have been handled and the way they're currently handled here in Azure.
Now, some preflight considerations that you really should go through before you deploy any workloads here in Azure. And the 1st 1 is going to be networking.
Networking is something that is a little bit hard to change, especially once you have your entire cloud network built out.
Your cloud network also needs to take into consideration your on premise network, especially if you're gonna be connecting those things up with a VP in. You need to make sure there are no sub net conflicts, no address conflicts, and a whole host of other considerations should be taken into view
before you launch your first virtual machine.
Some of these things you can change and undo here in Azure other things are more difficult to change just because of the nature of the way networking works. We'll talk about networking in the upcoming major section of the course, but definitely have in mind what the network is going to look like before you deploy your first production work, Lou.
In addition to that,
naming is another concept that people don't often think about, particularly as its changes things here in the cloud and some really great things to consider when it comes to naming our including things along the lines of
what environment your machine is gonna be in. So whether that's development, production or testing, you can include things like the virtual machine size or type. You can include the region that virtual machine is in, and you could also include a generic identify air. So if you have multiple machines that fit the same bill,
you can identify them one from the other.
Most companies have a naming schema. But I advise you thinking through what that looks like and how cloud considerations might change your current corporate naming ideas.
The next thing that we're going to discuss is the concept of regions. How that affects availability of service is, and the pricing of those service is not to start with what exactly is a region and a region is just a subset of a geography. And what is a geography? Well, a geography is exactly what it was when you were in school. So
North America miss a geography. South America is a geography,
and a region is going to be within one of those geography ease. And a region is simply a subdivision of data center or a location here in the azure environment where there is data center presence now, we'll talk later in this section of the course about availability zones
and availability. Zones are essentially
a second or third data center facility within the same region, and a region is going to be an area that, with multiple data centers
are connected by azure, is core backbone network, and so workloads can shift between those availability zones but still ultimately stay within that region. There are a number of regions here in the United States. For example, there are a number of them in the east East. One east, too.
There is South Central. There is central. There's west regions. There are also government regions for government customers as well.
Now, how does that concept of regions impact my service availability? Well,
the service is available in each given region. All depend on what Microsoft is ultimately able to offer
said another way. Not all azure service's are available in all regions, and that's a little bit of legwork that you'll need to do before you begin rolling out your entire workload. Set in azure are the workload is that you want to roll out, available in the regions that are closest to your facilities.
That has a big impact on performance and ultimate rollout
region. Also effects pricing. Different regions have different prices for the exact same products. That's also gonna depend on how much Amazon is paying. I'm sorry how much Azure is paying for data center space for bandwidth and a whole host of other concerns, so
you definitely need to figure out what region do you plan on deploying your workloads in
primarily that's gonna be based on which regions are closest to your facilities or your customers
and are the service is that you want to deploy available in that region
and are those service is available at a price point that warrants using the region. That's closest to you.
Now I move again deploying virtual machines. There are a whole host of virtual machine families, and these virtual machine families are just a group or a subset of virtual machines with varying different specs that largely are configured to accomplish similar goals. For example,
a general purpose family here is
generally balanced between CPU and memory. These are gonna be great for testing and development, small to medium databases and low to medium Web servers, for example.
There are also compute optimized virtual machines, and those are gonna be just what they sound like. They're gonna be more heavily leaning towards the CPU side. Then to the memory side,
there are memory optimized
virtual machine groups,
storage optimized, and these are gonna be designed tohave Hi, disk through put. These are great for virtual machines running databases.
There are also GPU optimized virtual machines clearly gonna be good for rendering or modeling. Also great for machine learning.
And then there are also high performance. Compute virtual machines means you're going to be the most horsepower CPU machines out of the entire azure stack.
Now you can change your V m type or your VM family
at any given point in time. However, there are some considerations. When you do that, you need to be aware that re sizing of production BM will likely cause that machine to be automatically rebooted.
And that can cause some changes in configuration with things primarily the i P address, depending on how you have the I P address of that machine set up. So
be aware that you can resize your virtual machine, but it will require a little bit of downtime and is a task best reserved for after hours or off hours times.
Now, another concept that has changed relatively recently here
in the life of Azure is going to be the concept of virtual machine disks. Now surely everybody's familiar with a virtual machine and the disc required to run it. But here in Azure, we have a couple of interesting concepts, and the most recent change is going to be that between a managed versus an unmanaged disc
now previously with a classic deployment model that we've talked about disc for virtual machines were unmanaged. Essentially, you had to sit been up your very own storage account. Put your image of your virtual machine on to that storage account and you pointed your virtual machine to that storage. And it used that image to boot.
Thankfully, now with the Army P I. Microsoft has streamlined that process, and the concept of management disc has come to pass. A management disc is still rather, a managed disc is still storage that your virtual machine image lives on, but it is spun up as a part of
the virtual machine itself. You do not have to create it. In fact, you don't really even see it other than seeing it on the virtual machine as a C drive where the operating system is mounted.
Everything in the next concept of operating system disks. So going forward, we really recommend that you use managed discs. There's not a whole lot of downside at this point,
and the operating system disc is going to live on a managed disc, and that's just what it sounds like. That's where the operating system lives. The OS disc is going to be the C drive on the machine, and the maximum capacity of the operating system disc is gonna be 2048 gigabytes.
It will be a say to drive and again labeled si
another type of disorder going to be temporary desk. And these are gonna be used for Paige files and things of the sort. They are not resilient, so you may lose the data on them during a reboot.
For the most part, the data will maintain during a reboot. But during maintenance events and other things that might affect your virtual machine, those disks will be deleted. There is no guarantee that those disks will persist. The data that's on them
no, on a Windows machine. This for temporary disc is gonna be labeled as D, and it's gonna be used for Paige files
and on Lenox machines. It's gonna be S D B. And it's gonna be formatted and mounted to the Mount drive by the azure Lennox agent
as a note, don't store data on that disk. It is just temporary storage.
The third type of disc here is going to be a data disk, and this is the place that you store things that you want to keep. The data disk is also going to be a managed disk. It's attached by scuzzy, and its maximum file size is 4000 gigabytes. This disc can also be any letter that you choose for it to be,
and depending on the family or the size of the virtual machine that you choose
will dictate how many data disks you can ultimately attach.
So in this episode, we talked through relatively quickly what some of the preflight considerations were for launching our first virtual machine. And that was gonna be primarily considering how you want the network to be laid out and what you want the naming scheme toe look like. Now that we're considering some cloud workloads in our naming format.
We also talked about the concept of regions or geographic areas within azure,
and how those regions ultimately affect the availability of service is, and also the price of service is
we went over the huge multitude of virtual machine families and please understand that within each of those families there are multiple different spect out virtual machines. So your choice here in Azure is very extensive.
We also talked about the difference between managed an unmanaged discs and the three different types of disks that you will run into here in Azure.
Coming up next, we are finally going to be getting our hands into running some computer work loads here in Asher.
Thank you so much for joining me. I'm looking forward to the next episode.